Distributed Sparse Block Grids on GPUs

Publikation: Beitrag in Buch/Konferenzbericht/Sammelband/GutachtenBeitrag in KonferenzbandBeigetragenBegutachtung

Beitragende

Abstract

We present a design and implementation of distributed sparse block grids that transparently scale from a single CPU to multi-GPU clusters. We support dynamic sparse grids as, e.g., occur in computer graphics with complex deforming geometries and in multi-resolution numerical simulations. We present the data structures and algorithms of our approach, focusing on the optimizations required to render them computationally efficient on CPUs and GPUs alike. We provide a scalable implementation in the OpenFPM software library for HPC. We benchmark our implementation on up to 16 Nvidia GTX 1080 GPUs and up to 64 Nvidia A100 GPUs showing state-of-the-art scalability (68% to 96% parallel efficiency) on three benchmark problems. On a single GPU, our implementation is 14 to 140-fold faster than on a multi-core CPU.

Details

OriginalspracheEnglisch
TitelHigh Performance Computing
Redakteure/-innenBradford L. Chamberlain, Ana-Lucia Varbanescu, Hatem Ltaief, Piotr Luszczek
Herausgeber (Verlag)Springer, Berlin [u. a.]
Seiten272-290
Seitenumfang19
ISBN (Print)9783030787127
PublikationsstatusVeröffentlicht - 2021
Peer-Review-StatusJa

Publikationsreihe

ReiheLecture Notes in Computer Science, Volume 12728
ISSN0302-9743

Konferenz

Titel36th International Conference on High Performance Computing, ISC High Performance 2021
Dauer24 Juni - 2 Juli 2021
StadtVirtual, Online

Externe IDs

ORCID /0000-0003-4414-4340/work/142252156

Schlagworte

Schlagwörter

  • Block grid, CUDA, Distributed data, GPU, Sparse grid

Bibliotheksschlagworte